In the age of desktop software, Microsoft had the luxury of taking years to copy competitors. In the age of Web software, it's next to impossible to catch up. As customers use websites, they generate data which helps the site's creator improve it continuously. It's a topsy-turvy world reminiscent of David Brin's The Practice Effect, which Microsofties would do well to read. Spending nearly three years to implement even a bad idea like bribing users to use its search engine shows how badly ossified Redmond's software-development culture has become.

From Steve Ballmer on down, Microsoft executives have been hyping its cashback-search program as "disruptive," "innovative," "game-changing" — whatever the popular buzzword in Redmond is this week. It is none of these. Paying users for search is an old idea, one that Bill Gates broached in 2005 and one that even the likes of Jeff Bezos have tried without success. Never mind the history; search-engine expert Danny Sullivan finds the actual product frustrating. Even if it were well-executed, Microsoft's cashback program will not succeed at anything other than drawing highly motivated bargain hunters — exactly the kind of shoppers retailers are uninterested in reaching.

That Microsoft has staked so much on such a bad idea is a sign of its desperation, one in line with its botched bid for Yahoo. In its renewed talks with that company, Microsoft has pinned its hopes on somehow getting ahold of Yahoo's search business, thinking that this is somehow the key to success.

Microsoft and Yahoo both need to try something else besides beating Google at search. What they need to do isn't to make a search engine better than Google's; they need to make search irrelevant, as tiresome a metaphor for computing as Microsoft's desktop interface seems now.

In the world of The Practice Effect, new goods are nearly useless, while old goods are valuable, improving rather than wearing down with every use. That is a neat way of thinking of the world of the Web, where algorithms get better, not worse, over time.

Udi Manber, who headed Amazon.com's doomed effort to compete in the search-engine world before jumping to Google, recently wrote about Google's efforts to improve the quality of search. Manber attempts to be modest in the post, but he manages to make the job of keeping up with the complexity of Google's search queries sound almost, but not quite, impossible.

That's actually good for Microsoft, if only its executives were bright enough to realize it. With Google obsessed by this challenge, it could move on to other fields of endeavor. Search is not the be-all and end-all of the Web; finding is the beginning of a task, not the end of it. For Microsoft, search is over. It's time to go practice something else.